Nissan RB25 CAS, how can I help?

[winstonusmc]

My RB26 runs the Hitachi CAS with the plastic cover.My engine has had all types of CAS (Mits, metal Hitachi, plastic Hitachi) with the same stock ECU and ran fine. I read on the AEM website that the Hitachi and Mits CAS read different. http://forum.aempower.com/forum/index.php?topic=16575.0. It was saying about changing the crank trigger to rising edge on the Hitachi. I tried changing the Ignition Input Capture to rising edge and it didnt register any RPM on cranking.

[winstonusmc]

Is there any way to run the Hi res signal on JS10? Maybe mimic the circuit on the MS3x for PT4, 1k resistor and the two diodes? I am also trying to figure out how to bypass the VR circuit for the crank. I tried tying the TachSelect directly to TSel with no success.

[jsmcortina]

I'm considering re-writing these modes to use the high-res for sync only and use the low-res for running.

James

[winstonusmc]

Does this mean that the high res will only locate which low res slot is which? I would think it works like this:

120 degree slot goes low -> start count on the 1 degree slots
120 degree slot goes high -> end count on the 1 degree slot

This would tell the ECU what location the cam is in, either 24,20,16,12,8, or 4

Keep in mind, I havent checked it on the Mits CAS yet

[jsmcortina]

The main aim would be that it would work "the same" but not lose sync!

James

[Matt Cramer]

If the problem is mostly as RPM comes up, would it be possible to use the high res signal while cranking and below a certain threshold, like 1500 RPM? I've seen accounts of the 240SXs often "kicking back" while cranking because of not having enough resolution during rapid RPM changes while starting, and being able to crank on the high res signal should help with that.

[jsmcortina]

I've got some ideas to try for a new version of the decoder.

James

[winstonusmc]

New alpha?

[jsmcortina]

I doubt it will be as soon as alpha 18 though.

James

[winstonusmc]

Got the drift car running the Mits. CAS to rev over 4500 rpm today. It barely ran over idle with sync loss with the "Flip Polarity on hi-res tach" on "normal" but reved freely without sync loss on "inverted". Couldnt actually rev it beyond 4500 because its not tuned beyond that and it just richens out. I didnt have my wideband hooked up to adjust it either. More tuning later this week.

[jsmcortina]

I re-wrote the Nissan and Optispark modes to use the hi-res for slot measuring / phase detection and cranking spark only. I also applied a +/-1 tolerance to the count to allow for phase shifts. I've stim tested the RB25 and Nissan 16-12-8-4 modes and physically tested my Optispark.

I've sent out some test code to a couple of users.

James

[winstonusmc]

Its going to be a couple days before I can test this. I am pretty busy, maybe even till Saturday. I left my patch harness in the racecar and wont be able to go get it till then. I should have taken it because I have to add a fan circuit wire to it anyway.

Will the lack of using the 360 degree slots affect the resolution on timing during fast acceleration? Or will having the 6 slots be enough resolution for this?

[jsmcortina]

The existing code already does "1st deriv prediction" to handle that situation.

This was a pragmatic design change. Using the hi-res slots all the time is "ideal" but given the variances in user's hardware and also in CAS implementation, it appeared we were flogging a dead horse. Better to try a revised approach.

James

[winstonusmc]

Just tried the test code, works but......... the timing is off. It revs to 4500rpm before it just wont go anymore due to something. I did set it to 0 offset as per James's instructions. I dont have a timing light and the wideband is on the other car. It idled better, not sure if it was from the timing being off. It will be a couple days before I can get to the shop to get it timed. I knew the timing was off because I was able to build about .3bar of boost while free revving . Here are my logs:

2012-01-08_19.52.17.msl

(47.67 KiB) Downloaded 60 times

2012-01-08_19.50.56.csv

(376.95 KiB) Downloaded 45 times

[winstonusmc]

Got everything timed today, it was -16 degrees off on the trigger angle. It idles much smoother and CL idle actually works like it should. I have been using VEAL for a little and got it to rev to 5000rpm with no sync loss. It goes into boost well but needs some accel enrichment, as I have enabled it yet. I feel real good about it, just need to get rid of the nasty hole, it unloads in third

MultiMAP2.msq

(157.46 KiB) Downloaded 55 times

[jsmcortina]

Good news. I'll aim to get that "-16" built in to the code before alpha 18.

James

[winstonusmc]

Quick update. After some disappointment last night when our drift car wouldnt run right with the experimental code, I switched back to the normal apha17. Today we pulled the car to the track to get some tuning and discovered something interesting. The Mits CAS wouldnt rev past 6000rpm without losing sync, so I switched back to the new code. It still ran funny, but I starting to tweak the timing. It ran better. Hooked a timing light to it and the final setting is -25. The car ran great and tuning could commence. Funny thing is both the metal hitachi and Mits were the same. My R34 plastic Hitachi was -16, weird. All of those sensors work the same on my R34 stock ECU. I had the new programming bouncing off the rev limiter. It did lose sync but I didnt notice, it only happened coming off throttle.

Settings for the R32/33 CAS

[winstonusmc]

Dont know how or why or just a rectal cranial inversion, but the final offset angle is -25 for all. I dont know why it was different the other day when I timed my Skyline, but it is the same as the older CAS. -25 degrees is final. Sorry James for the mix up.

[jsmcortina]

At least that makes more sense with them the same.

James

[winstonusmc]

All in all its running good. I get a hesitation almost everytime when going into boost above 3000 rpm. I am posting up a pic of one hesitation and have noticed that the RPMdot is pretty messy, is this normal? I am just trying to figure out if the CAS programming is 100%. I may still go crank trigger anyway.

2012-01-28_19.00.26.msl

(264.85 KiB) Downloaded 44 times